Vulkan Continues To Show Its Gaming Strength On Low-End Hardware
Written by Michael Larabel in Linux Gaming on 26 January 2018. Page 1 of 1. 31 Comments

As we have shown in past benchmarks, while current generation Linux games with current Linux GPU drivers using the Vulkan API rather than OpenGL may not be significantly faster with higher-end hardware right now, the impact of this newer Khronos graphics API tends to be more profound on lower-end hardware, especially when it comes to lightening the load on the CPU. Following the recent Pentium vs. Ryzen 3 Linux gaming tests, I carried out some fresh benchmarks looking at OpenGL vs. Vulkan on the Ryzen 3 1200 quad-core CPU with NVIDIA and Radeon graphics.

This article offers a fresh look at the OpenGL vs. Vulkan performance on low-end hardware in the form of the Ryzen 3 1200 quad-core at 3.1GHz (3.4GHz boost) and using the Radeon RX 560 and GeForce GTX 1050 graphics cards. The latest Linux drivers from each vendor were in use: the NVIDIA 390.12 driver and on the Radeon side was Mesa 17.4-dev built against LLVM 7.0 SVN via the Padoka PPA> Linux 4.15 Git was in use for both driver configurations. The OpenGL vs. Vulkan performance was then compared in current comparable Linux games like Mad Max, Dawn of War III, Serious Sam 3: BFE, and The Talos Principle.

Not only are we comparing the raw frame-rate performance under these configurations, but via the Phoronix Test Suite the CPU usage was also automatically monitored and recorded during the automated benchmarking process.

Mad Max at 1080p with low quality settings saw the NVIDIA Vulkan performance with the GTX 1050 to be significantly better than the OpenGL performance. In some cases, quite significant. With the RX 560 performance on RADV, the performance still generally trailed the RadeonSI OpenGL performance but not by much. In the stronghold scene, the RX 560 found better performance with Vulkan.

The CPU usage was dramatically lower on both systems when using Vulkan rather than OpenGL. In fact, the peak CPU usage when using Vulkan on both graphics cards / drivers was just a little bit higher than the average CPU usage when using the OpenGL renderer. Quite a dramatic difference especially on a low-end quad-core Ryzen processor.

When running Mad Max with 1080p high quality settings, the Ryzen 3 system with these lower-end graphics cards had a difficult time keeping up with a 60 FPS average. Similar to the results at low quality settings, the RADV Vulkan performance on the RX 560 was generally just shy of the RadeonSI OpenGL performance while for all the scenes the NVIDIA Vulkan performance was much better than OpenGL.

Even with the higher quality settings, the CPU usage on both systems remained much lower when using the Vulkan renderer/driver than with OpenGL.

With Dawn of War III at low quality 1080p settings, these graphics cards on the Ryzen 3 system would have a tough time delivering playable results. The Vulkan vs. OpenGL performance was similar in this case.

But with the CPU usage metrics we see again that using Vulkan significantly frees up resources on the Ryzen 3 processor.

When pushing high quality settings to stress these lower-end graphics cards further, the GTX 1050 was now faster by a few frames with Vulkan while using RADV on the RX 560 was still slower than RadeonSI OpenGL.

Again, to no surprise, the CPU usage is noticeably lower when using Vulkan.

Serious Sam 3: BFE with Vulkan is significantly faster on both systems than with the OpenGL renderer.

The frame-rates were substantially higher with Serious Sam 3 BFE on Vulkan while the CPU usage this time was similar to the OpenGL results.

When stressing the hardware by pushing Serious Sam 3: BFE to ultra quality settings, the NVIDIA Vulkan performance was faster than OpenGL while the RX 560 favored the more mature RadeonSI OpenGL driver rather than RADV Vulkan.

The CPU usage was slightly higher this time with Vulkan than OpenGL.

Talos Principle ran into some strange behavior with the OpenGL performance (perhaps due to the only 2GB vRAM size on this GTX 1050?) but the Vulkan performance put it in line with the RX 560. With the RX 560, the Vulkan performance was faster than OpenGL.

The CPU usage with the NVIDIA Vulkan support was lower than OpenGL. When looking at the CPU usage during Radeon testing with The Talos Principle, the RADV driver usage led to a slightly higher average CPU use but a lower peak CPU utilization.

Lastly are some results from Feral's latest port, F1 2017. F1 2017 is Vulkan-only but I included some results for those curious about this racing game on Linux....

With ultra low settings at 1080p, the RX 560 was slightly faster than the GTX 1050. Both had similar minimum frame-rates.

The overall CPU usage was on average lower with the GTX 1050, but the RX 560 peak CPU usage was lower.

With high quality settings at 1080p, both GPUs and current drivers had an average FPS around 60 FPS and their minimum FPS around 50+ FPS.

The CPU usage with NVIDIA was again on average lower but its peak CPU usage was higher.

These results show the RADV Vulkan driver still have more room to advance and mature compared to RadeonSI, but at least it's continuing to beat the open-source OpenGL driver in some of the tested Linux games. With the mature NVIDIA OpenGL/Vulkan Linux driver support, the Vulkan performance on the GeForce GTX 1050 generally was much faster than using OpenGL. For both the Radeon and GeForce drivers, when using Vulkan the CPU usage on the Ryzen 3 tended to be noticeably lower than when running the Linux games with OpenGL.

About The Author
Author picture

Michael Larabel is the principal author of and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 10,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and automated benchmarking software. He can be followed via Twitter or contacted via

Related Articles
Trending Linux News